Irreal

Today’s post is a quick and simple tip from Zachary Kanfer on improving isearch. He likes the way isearch finds its results incrementally but is unhappy with the lack of context concerning the number of results found at each step.

It turns out that it’s easy to fix this. There is a configuration item, isearch-lazy-count, that when set to t will cause the count of prospective matches to be displayed in the minibuffer. There are a couple more configuration items that control the placement and format of the count. See Kanfer’s post for the details.

An alternative, that I use, is swiper-isearch. It also provides a count and displays the line containing each match in the minibuffer. I like it better and remapped Ctrl+s to it instead of isearch. Nonetheless, I have Super+s mapped to isearch so I can use it when I like. It’s worthwhile adding the prospective match count to isearch and it’s easy to do so I will probably add it to my configuration.

Lots of people recommend Evil mode or switching to a Dvorak keyboard because it allows for more efficient editing. Evil mode is probably more efficient and, more importantly, helps with RSI problems. As a long term Vi/Vim user, those bindings were burned into my muscle memory and when I switched to Emacs and decided to use the vanilla keybindings, it was hard to give them up. Still, the default Emacs binding are also pretty efficient and the ability to easily implement my own editing commands more than makes for the loss of composability. The bottom line is that I’m pretty much agnostic concerning Evil mode.

Then there’s Dvorak and similar keyboard layouts. After suffering the pain of learning to touch type with QWERTY, I have no desire to start again for a possible gain of a few words a minute. I’m pretty efficient with QWERTY and if I needed to up my speed, there are plenty of videos and tutorials to help with that. The bottom line is that I’m not going to adopt some new keyboard layout and the associated pain for a small, putative increase in typing speed.

That brings us to Devil Mode. The idea is to get rid of the modifier keys while avoiding modal editing. To do that, the comma key is used instead of the Ctrl key so that you would open a file with , x , f. There are, of course, ways of dealing with the Meta modifier as well.

Your first question is, undoubtedly, how do you type a comma? The common cases of a comma followed by a Space or Return work normally but you can use , , for the edge cases. Follow the link for all the details.

If you really really want to eliminate the modifier keys without embracing modal editing, Devil Mode may be worth checking out. For me, it seems more like moving to Dvorak: whatever gain you might get isn’t worth the pain.

Eli Zaretskii writes to say that the second pretest of Emacs 29.1 is available for testing. If you don’t mind living a bit on the edge, Zaretskii would be grateful for your testing and letting the developers know of any problems you find.

As always, thanks to Eli and the other developers for their hard work in readying this latest release for our use. They are truly heroes and we owe them all our gratitude and, should the opportunity arise, a beer.

Franco Pasut has an instructive post on how to diff files with Vim and Emacs. Both editors have very nice built-in functions to make diffing files easy. I’ve used them both and can attest to their utility.

Pasut’s post, Notes on resolving differences between two documents with the built-in resources of Vim and Emacs, isn’t a comparison and doesn’t make any value judgments. Rather, it merely shows how to use the two systems. I haven’t used Vimdiff in years but my memory is that it was easier to use than Emacs’ Ediff system.

Regardless of their relative merits, both are easy to use and very powerful. To see this, think about how you would reconcile two files with just the standard Diff program and some random editor. With Vimdiff and Ediff, you see the differences one by one and can choose which variant to keep. With Ediff, you can even edit the file directly.

If you’re not familiar with the two systems, take a look at Pasut’s post. It’s short and tells you more than enough to use either efficiently.